3D Slicer: A Free Open-Source Platform for Medical Image Computing And Software Tool Sharing
1
Fedorov ,
2
Fillion-Robin ,
2
Finet ,
1
Pujol ,
Andriy
Jean-Christophe
Julien
Sonia
1
2
3
4
1
Fiona Fennessy , Stephen Aylward , James V. Miller , Steve Pieper , Ron Kikinis
1Brigham
and Women’s Hospital, Harvard Medical School, Boston MA
2Kitware Inc., Clifton Park NY 3GE Global Research, Niskayuna NY 4Isomics Inc., Cambridge MA
Overview
3D Slicer is a multi-platform, free and open source software package for
visualization and medical image computing.
Quantitative Analysis
Multi-modality Visualization
A combined visualization of multiple imaging modalities and derived data can provide
clinician scientists with an integrated understanding of anatomy and pathology.
Segmentation is required for defining features of interest in imaging data for
quantification and analysis.
• manual contouring and editing
• region growing and level sets
• graph cuts with gesture support
• skull stripping and hierarchical brain
The software platform is community created for the purpose of subject specific
medical image analysis and visualization. Slicer includes support for:
• 2-, 3- and 4-D image visualization
(reformats, volume rendering)
• modality-independent (MR, CT,
PET, US and more)
• parameter maps and VOIs
• surface models & glyphs
• measurement tools & annotations
• charts
• Multi-modality imaging including, MRI,
CT, US, nuclear medicine, and microscopy
• Multi organ from head to toe
• Communication interface for external
devices (image-guided interventions)
• Expandable and interfaced to multiple
toolkits
• Installers available for Windows, Mac
and Linux
• Community, documentation & support
Slicer is intended for research work and has no FDA clearances or approvals of any kind. It is the
responsibility of the user to comply with all laws and regulations (and moral/ethical guidelines) when using
the software.
Sharing of Methods and Tools
3D Slicer consists of the core application and extension modules. The core contains
functionality of general applicability and is intended to be lean and robust. Domainspecific tools can be shared via open contribution process as extensions to the 3D
Slicer core.
Extensions statistics as of Feb 2013:
• Application store concept (similar to Apple
• 23 extensions
AppStore, Google Play or Chrome Extensions) • contributors from 6 institutions
• Curated by Slicer core developers
• Main categories: DTI processing, IGI,
• Same expectations apply to extensions as to quantitative image analysis, informatics
the main application:
• available for Windows, Mac & Linux
• wiki-style documentation
• Nightly and Stable builds
• same process of rigorous testing and
packaging
• Except: extensions are not required to
adopt Slicer license - license is chosen by
the extension contributor
Visualization of multi-parametric prostate MRI dataset that includes dynamic
contrast-enhanced (DCE) DWI and T2-weighted MRI. Interactive plotting of the
signal intensity over time is provided by Slicer MultiVolume support
infrastructure. DCE MRI pharmacokinetic modeling functionality is currently
under development and will be available as Slicer extension.
Examples of automated 3D segmentation results obtained using GrowCut (top) and
FastMarching effects of the Editor module. Initialized with few strokes (left), these tools can
produce meaningful segmentation results within minutes of computation.
DICOM module (left) provides
support for import, query, retrieve
and storage of clinical images
using DICOM protocols and data
structures. Domain-specific parts
of DICOM are supported via
extensions.
Visualization of 3D reconstructed transrectal ultrasound of the
prostate. Reconstruction performed using Public Library for
Ultrasound imaging (PLUS) (PerkLab) http://plustoolkit.org
Time-series and multi-subject analysis
require registration of imaging data
acquired at different times, on different
scanners, and across modalities.
Pre-treatment
Supported registration types:
• rigid, affine, non-rigid, fluid transformation
models
• fiducial-, surface-, intensity-based algorithms
• scalar, vector, tensor image types
Extensive support is provided for various
modes of volume rendering. Complex 3D
visualization scenes can also include
rendering of surface models, image volume
reformats and annotations.
Slicer was initiated as a masters thesis project between the Surgical Planning Laboratory at the Brigham
and Women's Hospital and the MIT Artificial Intelligence Laboratory in 1998. Over the last decade, Slicer has
been supported by an active community of academic and research partners, enabled by the funding from
the National Institutes of Health.
Slicer License: Slicer binaries and source code are available under a BSD-style, free open source licensing
agreement under which there are no reciprocity requirements, no restrictions on use, and no guarantees of
performance. Slicer leverages a variety of toolkits and software methodologies that have been labeled the
NA-MIC kit, including ITK, VTK and CMake.
segmentation for morphological studies
• interactive visualization of the result
Visualization of a multi-modal brain tumor resection planning dataset. 3D
visualization includes surface model of the tumor and ventricles boundary,
fiber tracks, and fiducial annotations.
Numerous research formats
(such as NIFTI and NRRD) are
also supported. Initial support of
Annotation Image Markup (AIM)
is provided in Reporting
extension.
non-registered
Support of community standards and
interoperable formats is necessary to
enable communication with commercial
systems and collaborative research.
Community, Learning and Support
Contributing Extensions
Comprehensive infrastructure is available to support extensions developers in
contributing new functionality to 3D Slicer. Extensions can be developed using C++ or
Python languages. Various levels of integration (batch mode executables vs interactive
tools) are supported.
To support user and developer communities, and to facilitate effective translation of
tools into the clinical research setting, the 3D Slicer Project provides many outreach
materials and activities.
• Hands-on Training Workshops
• 2012: at 17 national and international
Key infrastructure components:
• CDash: continuous and nightly testing and
packaging of Slicer application and extensions
• Midas Platform: organization and storage of the
test data, cross-platform packaged application and
extensions
• Slicer wiki: documentation, guides and tutorials
• github.com: web-based service for hosting of the
extensions index and bug tracking
A
B
The process of contributing Slicer extension consists of the following steps: (1) extension
is developed following Slicer developer instructions; (2) the source is published under
open source or restricted license; (3) documentation is developed to present the
extension to the developer and user communities; (4) once extension is vetted by the
Slicer community, extension developer contributes extension description file to Slicer
extension index hosted on github; (5) extension is built, tested and packaged for all
platforms by CDash; (6) packaged extensions are uploaded to Midas Extension server
and become available to the user in the Nightly Slicer build Extension Manager.
Slicer extensions are installed and downloaded by the user from Slicer Extension Manager,
which includes basic information about the extension (category, description and
screenshots). Feedback/rating mechanism is currently under development. Longitudinal
PET/CT analysis extension (shown) integrates ROI segmentation, PET Standard Uptake
Value (SUV) quantification and versatile visualization of multiple time-points into a single
processing workflow.
CDash implements cross-platform testing and packaging of the main application and
extensions, with the web-based interface to report any issues during build or testing
process. Separate packages are created for the stable and nightly installers. Continuous
builds are triggered by updates to Slicer ExtensionIndex and can be used to quickly
identify and fix the problems before the nightly build is packaged.
Post-treatment
Example of using 3D Slicer BRAINSFit module to compensate for breast deformation
and position changes before and after treatment by means of deformable registration.
Mutual Information intensity-based registration; transformation models included affine
and B-spline transformations.
Summary of 3D Slicer version 4 download
statistics over 2012. Overall, the application
has been downloaded more than 48,000
times, with Windows being the most
popular platform. Most of the downloads
were initiated by the users from outside the
USA.
D
Screenshot of the Dose Volume Histogram (DVH) module developed by SparKit project and
available in SlicerRT extension. The DVH module compute metrics (left) and DVH curves
(top-right). Input dose, CT, and structure set data is displayed in the 3D viewer (top-center)
and slice viewers (bottom-right). SparKit is a project funded by An Applied Cancer Research
Unit of Cancer Care Ontario with funds provided by the Ministry of Health and Long-Term Care
and the Ontario Consortium for Adaptive Interventions in Radiation Oncology (OCAIRO) to
provide free, open-source toolset for radiotherapy and related image-guided interventions.
registered
Three levels of extensions compliance:
• Level 1: Fully compliant: Slicer license, open
source, maintained.
• Level 2: Open source, contact exists.
• Level 3: All other extensions (work in progress,
beta, closed source etc)
Project week events for researchers, Slicer users and developers take place twice a year and are open for anyone
to join. Training workshops are organized at major venues, such as RSNA, ISBI, MICCAI and SPIE annual
meetings.
venues
• Tutorial Materials & datasets
• Reference Style Documentation
• video tutorials and webcasts
• Active user and developer mailing lists
• 2012: developer and user lists: 507
and 869 subscribers; 1,701 and 3,350
messages
• Weekly developer meetings open to
everyone (Google Hangouts)
• Project week events
• Active user community (>48,000
downloads per year)
Acknowledgments 3D Slicer
Project is made possible with
partial support from NIH through
the following grants:
• U54 EB005149 (NA-MIC)
• P41 RR013218 (NAC)
• P41 EB015898 (NCIGT)
• R01 CA111288 (BRP)
• U01 CA151261 (BWH-QIN)